1. Field of the Invention
The invention relates generally to rocket engines and, in particular, to an aerospike rocket engine.
2. Background Art
Conventional rocket engines use round, bell-shaped nozzles. These nozzles, however, have an inherent limitation in that the combustion gas, or plume inside the nozzle can expand only as far as the shape and length of the nozzle allow, resulting in substantial under and/or over expansion, with a resulting loss of thrust and instability/vibration of the expanding plume. Bell nozzles are, therefore, typically designed for specific applications, e.g., take-off, high altitude, or outer space. However, even within the confines of these applications, under/over expansion invariably occurs due to 1) changes in atmospheric pressure, and 2) a finite expansion capability of approximately 1:400 (where infinite expansion is theoretically required in space), which may result in up to 5% loss of thrust. See, e.g., Missile Engineering Handbook, van Nostrand, Fig. 7.1.1, 1957; Aviation Week & Space Technology, p. 130, Aug. 10, 1987). Therefore, a bell nozzle having a given size and shape can reach peak efficiency only at an altitude where the plume expansion within the nozzle equals the theoretical expansion that would be permitted by the atmospheric pressure at that altitude.
To overcome the bell nozzle's limitation, Rocketdyne Propulsion and Power (“Rocketdyne”), a subsidiary of the Boeing Co., developed a nozzle which resembles a bell nozzle turned inside-out called an “aerospike” nozzle. More specifically, a linearized version of the aerospike nozzle called a “linear” aerospike nozzle was developed for the proposed X33/VentureStar single-stage-to-orbit (“SSTO”) space plane project. The linear aerospike engine resembles a bell-shaped nozzle that has been split in half and the two halves put back-to-back to each other, and the end of nozzle clipped or truncated. In some cases, however, the linear aerospike engine may have only one of the two halves, i.e., a single-sided engine. Because the plume of the aerospike nozzle is manifested on the peripheral of the nozzle, it is free to expand, limited only by atmospheric pressure. As a rocket using the aerospike nozzle climbs higher and higher, the plume is able to expand continuously against the decreasing atmospheric pressure, albeit at a cost to the thrust vector which diverges progressively sideways.
Referring to FIG. 1, there is shown a bank of five linear aerospike engines 10 arranged side-by-side. Each aerospike engine 10 comprises a rectangular wedge or tapered body 12, a slanted or curved reaction surface or plane 14, a leading end 16 and a trailing end 18. Each engine 10 has at least one injector 20 or, more typically, a set of injectors 20 adjacent the leading end 16 and arranged to direct a propellant or fuel down the reaction plane 14 towards the trailing end 18. Upon combustion of the propellant or fuel from the injector 20, the combustion gas, or plume, travels down the reaction plane 14 and exerts propulsive pressure on the reaction plane 14, which provides the thrust for the space plane.
As can be seen, turning to FIGS. 2A-2C, the linear aerospike design allows the plume to expand freely against atmospheric pressure. At low altitude, the exhaust plume 24 is held in a fairly narrow band 26 by the high atmospheric pressure as shown in FIG. 2A. However, referring to FIG. 2B, at high altitude and low atmospheric pressure, the plume 24 is able to expand. Shock waves produced by the supersonic speed of the space plane at high altitude provides a shock front 28 that can assists in resisting the expansion of the plume 24. As the space plane 22 climbs into outer space, the vacuum of space may tend to pull the plume 24 away from the reaction plane 14, as shown in FIG. 2C. This may result in “divergence,” wherein the plume's 24 thrust vectors becomes misaligned with the direction of flight, resulting in a decrease in net thrust and, hence, engine efficiency.
One solution to this divergence syndrome is to extend the reaction plane 14 so as to facilitate full expansion of the plume 24. However, because the plume 24 is unconfined, the boundary layer may tend to separate from the reaction plane 14. Boundary layer separation is a lifting off or peeling away of the plume 24 from the reaction plane 14. According to Bernoulli's law, as long as the boundary layer remains sufficiently energized, the plume 24 will adhere to the reaction plane 14 by virtue of the negative pressure between the high-speed boundary layer and the reaction plane 14. As the plume 24 travels along the reaction plane 14, the boundary layer may run out of energy and separate from the reaction plane 14. The effects of boundary layer separation include instability or turbulence which can produce severe mechanical vibrations that can damage the space plane 22. In addition, boundary layer separation may result in a loss of thrust and engine efficiency. Separation usually starts at the end of the boundary layer where the energy of the boundary layer is low. Atmospheric pressure can help to hold the plume 24 against the reaction plane 14. Therefore, separation is more likely to occur at high altitude where the atmospheric pressure is low.
One way of preventing boundary layer separation is by truncating the reaction plane 14 so that the reaction plane 14 is shorter (as can be seen in published illustrations of the X33). This allows the boundary layer to traverse the entire length of the reaction plane 14 before running out of energy. The trade-off, however, is that there is a reduction in thrust and engine efficiency relative to an untruncated reaction plane due to 1) under expansion, and 2) thrust vector diversion/deflection. Furthermore, the shorter reaction plane 14 may not allow the propellant or fuel sufficient time to completely combust/accelerate before reaching the end of the reaction plane 14, which can result in reduced thrust on the reaction plane 14. This reduction in thrust may be critical at high altitudes where the space plane needs to attain very high velocity.
Over and above the truncation limitation of the X33 implementation of the aerospike engine, scaling up of the aerospike plan form to suit larger space plane applications (e.g., the proposed VentureStar heavy lift shuttle) may additionally require cascaded or staged propellant/fuel injection in lieu of the impact of dimensional scaling.